Minimizing Dependency Length

Here I’m plotting data from the counterfactual languages optimized to minimize dependency length.

‘DistanceWeight’ is the distance parameter in the ordering model. Higher values indicate greater distance between head and dependent.

Distance logits for Det, Num, Adj

More negative values indicate stronger preference to be close to the noun. Here, adjectives are predicted to occur at higher distance from the noun than determiners.

Distance logits for Core Arguments

More negative values indicate stronger preference to be close to the noun. Subjects are predicted to come closer.

Distance logits for Nominal and Pronominal Arguments

More negative values indicate stronger preference to be close to the noun. Pronominal arguments are predicted to come closer than nominal arguments.

Values for other Dependencies

Here I’m plotting dependent-first and distance logits obtained by optimizing for dependency length.

For the “Distance” logits, higher values mean greater distance from the head.

For the “DH” (Dependent-First) logits, higher values mean greater preference for the dependent to come first.

As the number of dependency triples is large, I’m plotting averages along the dimensions head POS, dependency label, and dependent POS. I start with by-dependent averages since they seem easiest to interpret.

Averaged across Dependent POS

Here I’m plotting averages of inferred values by the POS of the dependent of each dependency triple.

Averaged Distance Predictions across Dependents

averaged over languages

Averaged Order Predictions across Dependents

Averaged across Head POS

Here I’m plotting averages of inferred values, by the POS of the head of each dependency triple.

Averaged Distance Predictions across Heads

There doesn’t seem to be anything to see here.

## Warning: Removed 1 rows containing missing values (geom_errorbar).

averaged over languages

Averaged Order Predictions across Heads

## Warning: Removed 1 rows containing missing values (geom_errorbar).

Averaged across Dependency Label

Here I’m plotting averages of inferred values by the label of the dependency.

Averaged Distance Predictions across Dependencies

## Warning: Removed 3 rows containing missing values (geom_errorbar).

## Warning: Removed 3 rows containing missing values (geom_errorbar).

## Warning: Removed 3 rows containing missing values (geom_errorbar).

## Warning: Removed 2 rows containing missing values (geom_errorbar).

## Warning: Removed 1 rows containing missing values (geom_errorbar).

## Warning: Removed 4 rows containing missing values (geom_errorbar).

## Warning: Removed 3 rows containing missing values (geom_errorbar).

## Warning: Removed 1 rows containing missing values (geom_errorbar).

## Warning: Removed 7 rows containing missing values (geom_errorbar).

## Warning: Removed 6 rows containing missing values (geom_errorbar).

## Warning: Removed 1 rows containing missing values (geom_errorbar).

## Warning: Removed 2 rows containing missing values (geom_errorbar).

averaged across languages

Averaged Order Predictions across Dependencies

## Warning: Removed 3 rows containing missing values (geom_errorbar).

## Warning: Removed 1 rows containing missing values (geom_errorbar).

## Warning: Removed 1 rows containing missing values (geom_errorbar).

averaged over Languages